Modeling and classification of service behaviors

With the drastic increase in the number of services, there is an urgent need to devise techniques that facilitate services selection. The behavior of a service is a key factor in such selection. One of the major challenges in this regard is to be able to model and recognize such behavior, especially when the service is a black box (i.e. no architectural details are provided). In this paper, we propose a new approach for modeling and classification of service behaviors. The proposed approach captures service performance through some predefined behavioral patterns. Each pattern is a typical sequence of observations in which an observation denotes the quality of a service for one interaction. We then follow a rough set based approach for the classification of services into different patterns. To show the applicability of the proposed approach, a comparative study with existing rule-based classification algorithms is provided.